GEOL 104 Dinosaurs: A Natural History

Fall Semester 2023
The Cretaceous-Paleogene Extinction: All Good Things...

Detail from "Dead T. rex" by Douglas Henderson (1999)

Key Points:
•The disappearance of non-avian dinosaurs was just one part of a larger event: the Cretaceous-Paleogene (K/Pg) mass extinction (formerly called the Cretaceous-Tertiary or K/T extinction).
•Diverse groups of land and sea organisms died out at this time, 66.05 million years ago.
•Three major environmental changes (the Maastrichtian Regression; the Deccan Traps volcanism; and the Chicxulub asteroid impact) occurred in this same general time. While the first two seem to have generated some level of extinction, it is the impact that seems to be the primary driver of this extinction.
•Survivorship on land seems to be controlled mostly by favoring small body size and ability to feed from stored food resources; in the sea it is the latter which is the most important factor.
•The western North American (Laramidian) fossil record of the last 15 million years of the Cretaceous is the most complete in the run up to the extinction event; there are some losses of groups before the impact that seem to have been driven by other environmental factor. That said, these do NOT seem to have been the cause of the loss of the non-avian dinosaurs.

Definitions and Dramatis Personae

One of the most interesting aspects of dinosaur history is the extinction of the non-avian dinosaurs and the rise of mammals to dominance.

How did this occur?

First, some definitions:
Extinction: not recognized as a natural phenomenon until Cuvier. Different definitions (or at least different emphases) according to different types of scientists:

(All of these essentially mean the same thing: there are no more of that kind of organism).

Only two types of taxa can go extinct: species and clades. Old-fashioned gradistic paraphyletic groups could "go extinct" even though their descendents (and thus their genome) persisted on.

So, in this sense, Dinosauria (and Saurischia [or Ornithoscelida], and Theropoda, and Neotheropoda, and Averostra, and Tetanurae, and Avetheropoda, and Coelurosauria, and Tyrannoraptora, and Maniraptoriformes, and Maniraptora, and Metornithes, and Pennaraptora, and Eumaniraptora, and Avialae, and Pygostylia, and Ornithothoraces, and Euornithes, and Carinatae, and Ornithurae) are not extinct!

Extinctions happen throughout the fossil record.

What interests us is Mass Extinction:

The effect of mass extinctions observed by William "Strata" Smith and others: the reason for dividing the Geologic Column into Eras and Periods is because of mass extinctions:

The end of the Mesozoic is the boundary between the Mesozoic Era and Cenozoic Era, which is also the boundary between the Cretaceous Period (K) and the Paleogene Period (Pg):

What died out?
Among the marine invertebrates:

Among marine vertebrates, taxa that died out at K/Pg include:

(NOTE: Some popular and scientific books and articles show ichthyosaurs as victims of the K/Pg extinction. However, this clade was extinct TENS OF MILLIONS OF YEARS prior to the K/Pg boundary)

Marine turtles survived the event, although some groups of marine turtles died out during Cenozoic. Also, one group of marine crocs (the dyrosaurids) survived the extinction event, only to die out early in the Cenozoic.

On land, victims include:

NOTE: Other groups of dinosaurs (e.g., other sauropod clades, stegosaurs, coelophysids, megalosauroids, carnosaurs, etc.) were already LONG EXTINCT and were thus NOT victims of the K/Pg event.

Important to remember that there were LOTS of survivors (otherwise, there would be no life today!!):
On land:

Looking for Causal Agents

When talking about "causes" of mass extinction, we have to recognize we are talking about several different things:

Species/Area Relationships and Extinction: Ecologists have observed, initially on islands but now confirmed for any contiguous region, that the number of species in a region scales closely to the area of that region. The larger the area, the more species (whether we are looking at plants, insects, birds, fungi, or whatever.) This region might be the continental shelf around a continent, a forest, an island, or whatever.

What happens when the region shrinks? For instance, when a forest is cut down, or sea levels drop and the continental shelf is reduced? The number of species that habitat supports goes down. In other words, some of these species are going extinct, at least locally. If that region was the entire home range of that species, that means it is the true extinction (rather than just a local extirpation.) Thus, environmental changes that reduce habitats are likely to effect many species across that region (hey, sounds like turnover-pulse, doesn't it? Hmmm....)

Now what happens if the habitat increases again? Species don't just magically appear to fill it up. If the expanded habitat recontacts similar habitats, they might share their species and re-equilibrate to the higher level. Otherwise, they will need to wait for new species to evolve (or migrate in from some other region) until equilibrium is reached again.

Many hypotheses proposed for the K/Pg Extinction. In evaluating the hypotheses, must consider:

Most significantly: (Without those two aspects, the hypothesis is not scientific, but simply speculation)

Here are but some older proposed causes for the K/Pg event:

  • Racial Senescence
  • Poison Gas from Comets
  • Caterpillars ate all the food
  • Mammals ate the dinosaurs to death
  • Allergies to Angiosperms?
  • Diseases

    Will explore other hypotheses...

    Modern Approaches to the Cretaceous-Paleogene Extinction

    The global nature of the K/Pg extinction would seem to favor some causal agent which could affect the whole planet. Cosmic (extraterrestrial) phenomena might be a good possibility.

    1971: Suggestion by Dale Russell (dino paleontologist) and Wallace Tucker (astrophysicist): a supernova killed the dinosaurs.

    Supernovae are exploding stars: put out TREMENDOUS amount of energy. If a star in a nearby solar system exploded, it would bombard surface of planet with radiation, bringing radiation sickness, cancer, etc.

    Modern analogue: during 1950s through 1970s, greatest fear about nuclear war was radioactive fallout.


    Fits prediction. However, problem because it is an untestable (and thus non-falsifiable) hypothesis:

    So, remains as a potential but no reason should be supported. Was the leading candidate during the 1970s.

    Confirmed Potential K/Pg Causal Agents

    While the above phenomena are largely insufficient to explain the event, there are three confirmed large-scale environmental changes during the end of the Maastrichtian that potentially are involved with the extinction. In order of appearance (and of increasing severity) they are:

    Maastrichtian Regression:
    The Maastrichtian is last Age of Late Cretaceous Epoch; regression refers in this context to any period of sea-level drop. In this particular case, it was triggered by reduction in mid-ocean ridge activity. As the mid ocean ridges shrank, the water that they had displaced onto the continents throughout the Late Cretaceous drained away.

    For the marine realm, this caused a very large drop in the area for the shallow marine communities. Additionally, it removed a major source for marine productivity and modified global circulation patterns. For the terrestrial realm it made climates more continental (hotter summers, colder winters) because now not every place was essentially close to the sea shore. This would also mean that habitats would shift.

    The Maastrichtian regression unquestionably occurred (although as with all such changes, there were local and global smaller scale sea level rises mixed in with the pattern of the drop.) One would predict that the environmental effect of the Maastrichtian regression would play out over the period of a few million years.

    Deccan Traps Volcanism:
    The Maastrichtian has long been known as a period of intense volcanism in parts of the world (see "Global Diastrophism" above). For instance, in North America, associated with change in mountain building in Rockies (the beginnings of the Laramide Orogeny). But the biggest aspect of this volcanism is the Deccan Traps.

    The Deccan Traps were a GIGANTIC series of lava flows in western India. They were the most major flood basalt event since the Siberian Traps at the Permo-Triassic boundary and the Central Atlantic Magmatic Province volcanism that formed by the break up of Pangaea at the Triassic-Jurassic boundary. In some places the Deccan Traps are 2.4 km (1.44 MILES) thick. An area of 2 million km3 was covered. As with all flood basalts, it was not a single continuous event; instead, there would be eruptions; periods of cooling from sulfates; global warming from excess greenhouse gases; stabilization of the environment after the eruption had ended; then a new eruption.

    The Deccan Traps began around half a million years before the K/Pg boundary. The lowermost beds erupted during a global magnetic normal time (subchron C30n), while the impact of the asteroid and the boundary itself would be later in a magnetic reversed interval (subchron C29r). Since the switch from C30n to C29r happened 350 kyr prior to the boundary, the asteroid impact could NOT have been the cause of the Deccan Traps. That said, there is some evidence that a few thousand years after the impact the intensity of the eruptions increased: this timing would be consistent with magma chambers being disrupted by impact-generated mega-earthquakes, and then percolating slowly up to the surface.

    As with the Siberian Traps and CAMP, the Deccan Traps would provide both a cooling and a warming component. There is a 5-6°C increase in annual temperature in western North America during the pre-impact eruptive phase, that might reflect the volcanic greenhouse contributions. Furthermore, we would expect ocean acidification as a side effect.

    So, Deccan Traps themselves were a MAJOR event, and might have contributed to the extinction event. Had just this eruption plus the Maastrichtian regression occurred, there probably would have been a mass extinction. But perhaps it would have been much less severe, and not an "era-ending" one. But Nature had one more, far more spectacular event to unleash.

    The Chicxulub Impact:
    The discovery of an asteroid impact at the end of the Cretaceous began in the latest 1970s and was published in 1980. Geologist Walter Alvarez was investigating a layer of clay in Gubbio, Italy at the K/Pg boundary. He wanted to determine length of time represented by the clay layer. He knew that the limestone below was latest Maatrichtian and those above were earliest Danian (early Paleogene). But how much time was there for the clay layer? A few years? Decades? Millennia? More? (The biostratigraphy above and below showed it couldn't be more than a million years.) Being clay he couldn't radiometrically date it, and there is no magnetic flip-flop right near the boundary. So he consulted his dad (Nobel winning physicist Luis Alvarez) for possible solution. After teaming up with a few chemists, the decided to look for meteoritic material as a possible clock. This was based on the following observations:

    1. Meteors impact the Earth's atmosphere all the time
    2. Some chemical elements more common in meteors and such than on Earth's surface: these should be traceable in minute quantities in sediment
    3. Find the average infalling rate of these elements today; use this rate and observed amount at the Gubbio clay layer to find out how much time

    The element used: iridium (a platinum-like metal, common in metallic asteroids but very rare in Earth's crust).

    When examined Gubbio clay, found a huge increase in iridium (the iridium spike) at base of clay. Using their method it would mean that there was a few million years missing, which was clearly not the case. So which assumption was not correct? They eventually realized that this wasn't a case of an average rate of infall. Instead, they hypothesized that an asteroid impacted Earth at the K/Pg boundary. Luis Alvarez had been part of the US nuclear weapons program, and later had worked with NASA on calculating the energy of impacts from asteroids and comets. Both of these sets of skills proved useful in determining the effect of the end-Cretaceous impactor

    The iridium spike was subsequently confirmed at many sites across the planet. The Alvarez team recognized that the primary killing agent wouldn't be the blast as such, but rather the debris kicked up into the atmosphere. Some of this would rain down right away, yielding a thermal pulse. The finer particles would remain suspended in the stratosphere, blocking out insolation and plunging the world into an impact winter of frozen temperatures and mass starvation.

    Although there had been a very good chance that the impact had hit oceanic crust that had since been subducted back into the mantle, the crater was actually recognized in 1990s. It is in the Yucatán Peninsula of Mexico. It is not visible from the surface except by radar scanning, because it is covered by 300-1000 m of Cenozoic sediments. But cores drilled during petroleum exploration had found a disrupted and melted layer right around the K/Pg boundary; with the publishing of Alvarez paper and subsequent search for evidence to test their hypothesis, these cores were re-examined and re-interpreted as being from a crater. Seismic and gravity scanning revealed the presence of a 180-km diameter crater, exactly the right size for hypothesized impactor. There are even sinkholes (cenotes) in the rock above that trace out the disrupted layer, because groundwater in the limestone preferentially drains down into the crustal fractures. The crater is named Chicxulub after a town near the first core where the disruption was recognized.

    Between the initial Alvarez paper and the discovery of the crater, several proxies of the impact beyond the iridium spike were recognized. These include:

    The map of the K/Pg sites and the thickness of the disrupted layers form a bullseye around Chicxulub, showing that this rather than some other spot was the source of the blast.

    Effects of Chicxulub

    An asteroid impact is a very different kind of causal agent, and its effects are more instantaneous than the other types of mass extinction causes we've explored. (Some of the effects, however, are extended over long periods.) Here is what we have reconstructed so far.

    Phase I (the same day): Shockwave and Tsunami: A magnitude 10.1 earthquake would be expected from this blast. In the modern world this would likely topple buildings all over the planet; in a Cretaceous world it might be very disruptive to forests and cause landslides (and make dinosaurs fall over...), but wouldn't be a primary killing agent. A shockwave of hurricane-force winds would spread over southern North America and northern South America would also be locally dangerous, and the energy flash from the impact would incinerate everything in line-of-sight, but again these are not mass extinction-causers. A tsunami of 100-250 m would surge out, but again while regionally disruptive would not bring an era to an end.

    Phase II (later the same day): "Easy Bake Oven" and the Canopy Collapse: However, other events of that first day would be globally catastrophic. Infalling material would mostly burn up in the atmosphere. Your average meteor doesn't put out much heat, but so much infalling material generates substantial infrared radiation. This heat raises air temperature by only about 10C° (18F°), but would be fully absorbed by rock, leaf, flesh, and any other opaque material. It is predicted that the increase in infrared radiation would be 8-10x that of high noon at the hottest spot of the Earth, and persist for many minutes to hours. Living tissue would bake, unless underground 10 or more cm (heat wouldn't have time to make it that deeper) or underwater (upper few microns of water might boil off, but that would be it). I have nicknamed this the "Easy Bake Oven" effect, and may be the reason that no land animal larger than 5 kg seems to have survived.

    Related to this, some of the material WOULD make it down to surface, and seems to have sparked off global forest fires. This is called the canopy collapse, and is shown by an increase of soot and charcoal and by an increase of fern spores (the fern spike in sediments post-impact. (Ferns are excellent at recovering from periods of forest fires.)

    Phase III (the first decade or so): Impact Winter: Material vaporized by the impact is kicked up into the stratosphere, blocking insolation. This was the primary killing agent suggested by the Alvarez team. Probes on Mars show big temperature drops when fine particles are spread to the high atmosphere. And in human history, the eruption of Tambora in Indonesia in 1815 produced chilling effects worldwide for more than a year later; later eruptions, such as the 1991 eruption of Mount Pinatubo, while not as severe, were better studied and analyzed.

    Estimates of duration of the Impact Winter have varied from a year or so to a few months to just a few weeks, but a model published in January 2017 puts a duration for a 26°C (46.8°F) temperature drop in global surface temperature for 3-16 years and a greater than 30 year duration until recovery! Recovery on land goes much more quickly (basically once the skies are clear), but some deep-water sites show the cooling required 10 kyr to recover.

    Phase IV (100 kyr or so): Greenhouse Warming: The impact site was covered by carbonate rocks; when you oxidize carbonate rocks, you release CO2 (this is the same as when limestone is converted into cement). The Deccan Traps already released enough CO2 to raise the atmosphere from ~500 to 1400 ppm; the new addition brought it up to 2300 ppm or so. Once the dust was cleared the full effect of the greenhouse gases could be in play. Recent studies show a warming of 5C° in the shallow ocean for 100 kyr; air temperature warming might be between 4.3 and 13.5C°C (average of 7.5°C). See the discussions of the P/Tr and the PETM a few lectures ago about the effects of this on the living world.

    Tanis: A Remarkable Snapshot of the Impact's Effects: In 2019 a report was published of a phenomenal site from the very latest Cretaceous of North Dakota. Named "Tanis" (after the site of the Well of Souls in the classic Indiana Jones movie Raiders of the Lost Ark), this site shows the iridium and spherule layers like many other K/Pg localities. However, this site also has a number of other unusual features. The sediments below the boundary layer are consistent with a tsunami-like surge from the last remnant of the Western Interior seaway splashing up into the coastal river system. The amazingly well-preserved fish at the site have impact spherules in their throat and gills: this indicates they were swimming around and feeding as the first droplets of molten Mexico rained down onto North Dakota. A non-technical report suggests that there are many other significant fossils to be described from this locality. New data based on the isotopic evidence from the fish scales and the age of development of the insect larvae at this site suggest that the impact was in late Spring/early Summer in the Northern Hemisphere (i.e., around late May or early June).

    General Patterns of the K/Pg Impact

    In the marine realm, plankton (including the larval ammonoids) and nekton suffer worse than benthos. Photosynthesizers, and the creatures that feed on them directly (and those that fed on THEM directly) suffer worse than bottom feeders (which eat food "stored" in sediment). Groups with symbiotic algae also suffer strongly. Among marine vertebrates, larger ones in the open seas suffer worse than those on the coast, and those with higher metabolic rates worse than those with lower ones.

    (Note: the basic pattern is the exact opposite of the P/Tr, where the benthos got clobbered relative to nektonic and planktonic forms.)

    In the terrestrial realm, freshwater animals and those that feed on the water ecosystem tended to do better than those which fed on land. Larger animals and medium-sized animals with high metabolic rates suffer worse than small animals and medium-sized animals with low metabolic rates.

    For the terrestrial realm, it seems that "Easy Bake Oven" is the primary selective filter, with the Impact Winter taking out some survivors and the Greenhouse Summer even more. For the marine realm the "Easy Bake Oven" is likely not a factor, but the Impact Winter is the dominant selective force.

    The pattern of extinction looks like it was global and essentially instantaneous: hours to days to months to a few years. This suggests that the Chicxulub impact is by far the major causal agent. But all three events (Chicxulub impact, Deccan Traps volcanism, Maastrichtian Regression) are known to occur, so the earlier scenarios may have destabilized the ecosystems to some degree.

    All three events (Chicxulub impact, Deccan Traps volcanism, Maastrichtian Regression) are known to occur. Can we separate their effects in the geological record?

    Suggestions that all these systems were in effect:

    But, there are complications:

    And which, if any, seems to have an effect on dinosaur diversity?

    What does the dinosaur record show?
    Only a few spots on Earth have late Maastrichtian dinosaur record, and fewer still showing both the earlier Campanian (83.6-72.1 Ma) and complete Maastrichtian (72.1-66.0 Ma) record to see the change over the last several millions of years:

    Only good, continuous record from mid-Campanian through earliest Paleogene is western North America.

    The late Late Cretaceous dinosaur-bearing rocks of western North America spans the Campanian and Maastrichtian, and has earliest Paleogene rocks right above it. Sequences are known from Alberta and Saskatchewan down through Montana, Wyoming, and the Dakotas to Colorado and Utah, down into New Mexico. What does the dinosaur record of these units show us?

    Throughout these formations are the same basic groups of dinosaurs:

    Currently, changes in these are best seen in hadrosaurids and ceratopsids; definite changes in tyrannosaurids, ankylosaurids, and pachycephalosaurs; other taxa are currently too poorly sorted out at species level to be certain.

    Pattern among big ornithischians: short-snouted forms (centrosaurines, short-snouted hadrosaurines, lambeosaurines) die out earlier, while long-snouted forms (chasmosaurines, long-snouted hadrosaurines) remain common until the K/Pg boundary. This might reflect changing abundance of some form of vegetation, but that is not definite.

    This pattern consistent with long-term (millions of year scale) change associated with Maastrichtian Regression (and possible vegetation change). However, no evidence that latest Maastrichtian dinosaurs were declining WITHIN latest Maastrichtian: might well have continued on to live in post-Maastrichtian if not for Deccan Traps &/or Chicxulub impact.

    Furthermore, there is a general trend that the geologically-youngest genera (like Tyrannosaurus, Triceratops, Pachycephalosaurus, and Ankylosaurus, among others) were the largest representatives of their clades. So selection seemed to be favoring larger body size (probably due to competition). This is fine when times are good and food was plentiful. But when the world suddenly is freezing and starving, this is detrimental.

    It is noteworthy that non-dinosaurs had an attribute that made them potentially quite vulnerable to the changes of a mass extinction: their huge size range between babies and adults. Essentially all non-avian dinosaurs had significant ontogenetic niche shifts between hatchling to adulthood: each size acting as essentially a different species. (A side effect of this is that dinosaur ecosystems may have had fewer species in them than mammal ones, since each dinosaur species is ecologically the equivalent to many mammal species.) When the catastrophe came, each growth stage would have to be viable in the stressed time in order for the species to survive; if any size class wasn't supported, the species would go extinct. (The same would apply for animals which less difference between baby and adult, like birds and mammals. But in these cases, there would be fewer growth stages, and thus fewer potential "points of failure".)

    There are periodic papers that come out either arguing that dinosaurs were in a long-term decline prior to the extinction event or that they were doing fine until the very end. At the present time the sampling is too sparse to really make a strong conclusion (especially given the fact that only the Laramidian data is even vaguely close enough to be useful.)

    So, what caused the dinosaurs to die out?

    Three equally valid answers:

    Lessons Learned from the K/Pg Mass Extinction

    The fossil record is our only evidence of mass extinctions: the reality that when environments change too quickly and intensely that global ecosystems can collapse, taking millions of years to recover.

    But the K/Pg mass extinction in particular taught scientists two major lessons, often learned and studied by the same teams who initially examined the K/Pg event itself:

    So understanding the world-ending catastrophes of the ancient world can help us plan to prevent them in the future.

    A relevant videeo:

    To Next Lecture.
    To Previous Lecture.
    To Lecture Schedule.

    Last modified: 22 September 2023

    Detail from "The Cretaceous-Tertiary Impact" (2007) by Joe Tucciarone